The present invention relates to a network system comprising a plurality of network entities and a network management system for managing the network entities, and more particularly to a detection method of faults in the network entities.
One network system that is known has a plurality of network entities (hereafter also referred to as “NE”) and a network management system (hereafter also referred to as “NMS”) that manages the network entities. In such a network system, it is necessary to monitor and detect faults in the NEs. According to a conventional mode, system and method for monitoring and detecting the NE faults, the NMS applies polling to the NEs. It should be noted that the NE is also referred to as a network element, and the NMS is also referred to as a network monitoring device.
An example of a conventional monitoring method of network element faults is disclosed in Japanese Patent Application laid-open No. 10-308761 (hereafter referred to as a prior art document).
In the monitoring method of network element faults disclosed in this prior art document, the NMS applies polling to the NEs. Information on the adjacent NEs collected by the NEs and the state of connection ports are collected. Before the polling is applied to another NE, if the information collected from other adjacent NEs indicates that its NE direction is down, the polling is not performed and a load is reduced.
However, the prior art document shows the following problems.
The first problem is that, as the NMS applies polling to the NEs, an extra burden is caused if a processing load of the NMS is heavy for some reason.
The second problem is that the polling is not applied to the NEs that have faults, which delays the detection of the fault NEs when they recover from the fault.